How to Use the Unstructured MCP in OpenAI Agents SDK

Q: How can I use Unstructured data management with the OpenAI Agents SDK?

Your agent manages the entire lifecycle. You first use listdatasources to check inputs, then you run a workflow via triggerworkflowexecution, and finally, you verify where the results landed using listdatadestinations. The MCP Server handles all that orchestration for your AI client.

Q: Does the OpenAI Agents SDK handle job history when dealing with Unstructured data?

Yeah, absolutely. You use the listworkflowjobs tool to get a full audit log of every execution run. This means your agent can troubleshoot by looking at historical jobs rather than guessing what went wrong.

Q: What happens if I need to check data sources in Unstructured with the OpenAI Agents SDK?

You just call listdatasources. The MCP Server instantly provides a list of all configured remote connectors (S3, GCS). Your agent gets this information and can then pass it along for context or validation.

Q: How do I verify workflow details using Unstructured with the OpenAI Agents SDK?

You use getworkflowdetails. This function pulls up the exact configuration for any processing pipeline, so your agent knows precisely what steps are involved before it tries to run anything.

Q: Can I check data destinations when using Unstructured with the OpenAI Agents SDK?

Yep. The listdatadestinations tool shows you all configured endpoints, like Vector DBs or SQL connections. Your agent needs this list to make sure it can write the processed data where it's supposed to go.

Manage complex data pipelines with your AI client using the OpenAI Agents SDK.

See Vinkius in Action

Works with every AI agent you already use

…and any MCP-compatible client

MCP Servers - Free for Subscribers

Connect Unstructured MCP to OpenAI Agents SDK

Create your Vinkius account to connect Unstructured to OpenAI Agents SDK and route execution through our secure gateway. The platform manages server hosting, runtime updates, and security layers. Configuration requires no manual server provisioning.

GDPR Free for Subscribers

Setup Unstructured with OpenAI Agents SDK

Ask AI about this MCP

ChatGPT

Claude

Perplexity

Monitoring Job Status via MCP Server

The `list_workflow_jobs` tool lets your agent check active and historical job runs. You don't have to guess if a document pipeline is stuck; you just ask the OpenAI Agents SDK to list them, and it gives you the full history. This prevents wasted time debugging failed jobs. When an agent needs to know what ran five minutes ago or if a new process started successfully, it checks `list_workflow_jobs` via this MCP Server.

Mapping Data Connections with OpenAI Agents SDK

Need to know where your data comes from? Use the `list_data_sources` tool. Your agent calls this to see every connected remote location, whether it's S3 or Google Cloud Storage. It’s a quick way to verify connectivity. It gives you a definitive list of all available inputs for Unstructured data processing. This is key setup work before your specialized agents can even start running.

Setting Up Destinations with MCP Server

The `list_data_destinations` tool shows every place the processed data lands—think Vector DBs or SQL tables. The OpenAI Agents SDK uses this to map out your entire data flow, confirming where the AI-ready output will live. It’s about knowing the endpoint. Before an agent writes a single line of code that sends results somewhere, it runs this check on the MCP Server to ensure the destination is ready.

Setup guide

Set up Unstructured MCP in OpenAI Agents SDK

Prerequisites

Python 3.10+ installed
openai-agents package (pip install openai-agents)
Active Vinkius subscription with a valid endpoint token

1

Install the SDK

Run pip install openai-agents to install the OpenAI Agents SDK. The MCP integration is built-in — no extra dependencies needed.
2

Connect via SSE transport

Use MCPServerSse with your Vinkius endpoint URL. Replace [YOUR_TOKEN_HERE] with your token from cloud.vinkius.com. The SDK auto-discovers all Unstructured tools at runtime.
3

Create your Agent

Pass the MCP to Agent(mcp_servers=[server]). The agent receives Unstructured tools as native definitions — JSON schemas resolve automatically.
4

Run the agent

Call Runner.run(agent, prompt) to execute. The agent invokes the appropriate Unstructured tools and returns structured results. Copy the full example on the right to get started.

agent.py

import asyncio
from agents import Agent, Runner
from agents.mcp import MCPServerSse

async def main():
    async with MCPServerSse(
        url="https://edge.vinkius.com/[YOUR_TOKEN_HERE]/mcp"
    ) as server:
        agent = Agent(
            name="Unstructured Agent",
            instructions="You have access to Unstructured tools.",
            mcp_servers=[server],
        )
        result = await Runner.run(agent, "List recent transactions")
        print(result.final_output)

asyncio.run(main())

Get your connection token →

Independent Platform Disclaimer: Vinkius is an independent platform and is not affiliated with, endorsed by, sponsored by, verified by, or otherwise authorized by Unstructured. All third-party trademarks, logos, and brand names are the property of their respective owners. Their use on this website is strictly for informational purposes to identify service compatibility and interoperability.

Why Choose Vinkius

Vinkius connects your tools to AI with real-time monitoring and automatic cost savings — all from one dashboard.

Connect Unstructured now

Real-time monitoring

Live

visibility into every interaction

Connect your favorite tools to your AI and see exactly what's happening — every request, every response, in real time.

Built-in savings

60%

lower AI costs

Vinkius compresses data between your apps and your AI automatically. Lower bills every month — no configuration required.

Single dashboard

One

place for every integration

Every tool your AI connects to, managed from a single screen. One account, complete control.

Common questions about Unstructured MCP in OpenAI Agents SDK

Your agent manages the entire lifecycle. You first use `list_data_sources` to check inputs, then you run a workflow via `trigger_workflow_execution`, and finally, you verify where the results landed using `list_data_destinations`. The MCP Server handles all that orchestration for your AI client.

Yeah, absolutely. You use the `list_workflow_jobs` tool to get a full audit log of every execution run. This means your agent can troubleshoot by looking at historical jobs rather than guessing what went wrong.

You just call `list_data_sources`. The MCP Server instantly provides a list of all configured remote connectors (S3, GCS). Your agent gets this information and can then pass it along for context or validation.

You use `get_workflow_details`. This function pulls up the exact configuration for any processing pipeline, so your agent knows precisely what steps are involved before it tries to run anything.

Yep. The `list_data_destinations` tool shows you all configured endpoints, like Vector DBs or SQL connections. Your agent needs this list to make sure it can write the processed data where it's supposed to go.

Use it with your favorite AI tools

Connect this server to Cursor, Claude, VS Code, and more.

OpenAI Agents SDK sdk-python

Google ADK sdk-python

Pydantic AI sdk-python

Vercel AI SDK sdk-typescript